Skip to content


Switch branches/tags



Prospective environmental and economic life cycle assessment of vehicles made blazing fast.

A fully parameterized Python model developed by the Technology Assessment group of the Paul Scherrer Institut to perform life cycle assessments (LCA) of vehicles.

See the documentation for more detail, validation, etc.

See our examples notebook as well.

Table of Contents


What is Life Cycle Assessment?

Life Cycle Assessment (LCA) is a systematic way of accounting for environmental impacts along the relevant phases of the life of a product or service. Typically, the LCA of a passenger vehicle includes the raw material extraction, the manufacture of the vehicle, its distribution, use and maintenance, as well as its disposal. The compiled inventories of material and energy required along the life cycle of the vehicle is characterized against some impact categories (e.g., climate change).

In the research field of mobility, LCA is widely used to investigate the superiority of a technology over another one.

Why carculator?

carculator allows to:

  • produce life cycle assessment (LCA) results that include conventional midpoint impact assessment indicators as well cost indicators
  • carculator uses time- and energy scenario-differentiated background inventories for the future, based on outputs of Integrated Asessment Model REMIND.
  • calculate hot pollutant and noise emissions based on a specified driving cycle
  • produce error propagation analyzes (i.e., Monte Carlo) while preserving relations between inputs and outputs
  • control all the parameters sensitive to the foreground model (i.e., the vehicles) but also to the background model (i.e., supply of fuel, battery chemistry, etc.)
  • and easily export the vehicle models as inventories to be further imported in the Brightway2 LCA framework or the SimaPro LCA software.

carculator integrates well with the Brightway LCA framework.

carculator was built based on work described in Uncertain environmental footprint of current and future battery electric vehicles by Cox, et al (2018).


carculator is at an early stage of development and is subject to continuous change and improvement. Three ways of installing carculator are suggested.

We recommend the installation on Python 3.7 or above.

Installation of the latest version, using conda

conda install -c romainsacchi carculator

Installation of a stable release (1.3.1) from Pypi

pip install carculator


As a Python library

Calculate the fuel efficiency (or Tank to wheel energy requirement) in km/L of petrol-equivalent of current SUVs for the driving cycle WLTC 3.4 over 800 Monte Carlo iterations:

    from carculator import *
    import matplotlib.pyplot as plt
    cip = CarInputParameters()
    dcts, array = fill_xarray_from_input_parameters(cip)
    cm = CarModel(array, cycle='WLTC 3.4')
    TtW_energy = 1 / (cm.array.sel(size='SUV', year=2017, parameter='TtW energy') / 42000) # assuming 42 MJ/L petrol

    l_powertrains = TtW_energy.powertrain
    [plt.hist(e, bins=50, alpha=.8, label=e.powertrain.values) for e in TtW_energy]
    plt.xlabel('km/L petrol-equivalent')
    plt.ylabel('number of iterations')

MC results

Compare the carbon footprint of electric vehicles with that of rechargeable hybrid vehicles for different size categories today and in the future over 500 Monte Carlo iterations:

    from carculator import *
    cip = CarInputParameters()
    dcts, array = fill_xarray_from_input_parameters(cip)
    cm = CarModel(array, cycle='WLTC')
    scope = {
        'powertrain':['BEV', 'PHEV'],
    ic = InventoryCalculation(cm.array, scope=scope)

    results = ic.calculate_impacts()
    data_MC = results.sel(impact_category='climate change').sum(axis=3).to_dataframe('climate change')'seaborn')
    data_MC.unstack(level=[0,1,2]).boxplot(showfliers=False, figsize=(20,5))
    plt.ylabel('kg CO2-eq./vkm')

MC results

For more examples, see examples.

As a Web app

carculator has a graphical user interface for fast comparisons of vehicles.


Do not hesitate to contact the development team at



See contributing.


BSD-3-Clause. Copyright 2020 Paul Scherrer Institut.